Master data and product hierarchies as the backbone of FMCG reporting

Why master data is the backbone

Master data is the reference information that stays relatively stable while transactions flow around it: product identifiers, descriptions, hierarchies, pack sizes, supplier codes, store locations. In FMCG, the product master is the single most consequential dataset because almost every downstream metric (sales, share, distribution, forecast accuracy) is sliced by product hierarchy.

Get the product master wrong and every report built on top inherits the error. This is why data governancedata governanceData governance is the set of policies, roles, and processes that ensure data is accurate, secure, well-defined, and used responsibly across an organization.Voir la définition complète → teams at companies like Unilever, Nestle, or PepsiCo treat master data managementmaster data managementMaster Data Management (MDM) is the discipline of creating and maintaining a single, consistent, trusted version of an organization's core business entities like customers, products, and suppliers.Voir la définition complète → (MDMMDMMaster Data Management (MDM) is the discipline of creating and maintaining a single, consistent, trusted version of an organization's core business entities like customers, products, and suppliers.Voir la définition complète →) as a controls function, not an IT chore.

The core identifier: GTIN and GS1

The GTIN (Global Trade Item Number) is the barcode-level identifier for a specific product at a specific pack configuration. It is issued under standards from GS1, the global not-for-profit standards body (gs1.org) that also governs EAN and UPC barcodes used at point of sale.

Key mechanics to know:

• A GTIN is assigned per trade item, meaning per unique combination of brand, variant, pack size, and packaging. Change the pack size from 500ml to 750ml and you need a new GTIN.
• GTINs exist in a hierarchy: each selling unit (the single bottle) has its own GTIN, and the case or pallet it ships in has a separate, higher-level GTIN. Retailers scan the selling-unit GTIN at checkout; distribution centers scan the case-level GTIN.
• GS1 recommends (and many retailers mandate) a new GTIN whenever there is a material change to the product, in line with the GS1 General Specifications, the technical rulebook for allocation.

This is where reporting breaks. A "pack size change" promotion (say, temporarily +20% free) is technically a new product with a new GTIN. If a retailer's system maps that temporary GTIN to a separate SKU (stock keeping unit) row instead of rolling it into the base product's history, brand-level sales appear to dip even though nothing changed for the shopper.

SKU attribute trees: where markets diverge

A SKU is the retailer- or manufacturer-defined unit of inventory, built on top of the GTIN but wrapped in local attributes: category code, subcategory, segment, price tier, private-label flag, promotional status.

The problem: there is no single global master hierarchy that every retailer uses. Each grocer maintains its own category management structure, often five to seven levels deep, for example:

Walmart's structure, Tesco's structure, and Carrefour's structure will diverge in naming, depth, and grouping logic. CocaCocaCustomer Acquisition Cost: total sales and marketing spend divided by the number of new customers acquired over the same period.Voir la définition complète →-Cola Zero might sit under "Diet/Zero Sugar" in one retailer's tree and under "Cola" with a sub-flag in another. Roll up "Cola category sales" across both and you get different totals depending on which hierarchy performed the aggregation.

Manufacturers manage this with their own internal product hierarchy (often built in an MDMMDMMaster Data Management (MDM) is the discipline of creating and maintaining a single, consistent, trusted version of an organization's core business entities like customers, products, and suppliers.Voir la définition complète → platform such as SAP Master Data GovernanceData GovernanceData governance is the set of policies, roles, and processes that ensure data is accurate, secure, well-defined, and used responsibly across an organization.Voir la définition complète → or Informatica) and then cross-map it to each retailer's taxonomy, a manual, error-prone process usually owned by a category or trade insights team.

Global classification standards that help, partially

Two standards attempt to create common ground:

• GPC (Global Product Classification), maintained by GS1, offers a shared brick-level category structure. Adoption is inconsistent, retailers layer their own logic on top.
• GDSN (Global Data Synchronisation Network) lets manufacturers publish standardized product attributes (dimensions, ingredients, allergens) once, for retailers to pull, reducing manual data entry errors.

Neither standard eliminates the category hierarchy problem, because retailers still control how they group products for their own merchandising and reporting needs.

Where this shows up in your reporting stack

When you pull data from Nielsen IQ or Circana (formerly IRI), two major FMCG market measurement providers, they apply their own harmonized category definitions across retailers specifically to solve this problem, that is their core commercial value propositionvalue propositionA clear statement of the benefits your product delivers, the problems it solves and why customers should choose you over alternatives.Voir la définition complète →. When you pull data directly from a retailer's own point-of-sale extract or a EDI (Electronic Data Interchange) feed, you get their native hierarchy, unharmonized.

This explains a common real-world discrepancy: a brand manager sees "category share" numbers from Nielsen that do not match the retailer's own portal figures for the identical week and store set. Both can be correct. They are measuring different hierarchy definitions.

A worked example: reconciling two GTIN-to-SKU maps

Say Brand X sells a product with:

• Manufacturer internal SKU code: MFG-00456
• GTIN: 07613xxxxxxxx (14 digits, GS1 standard)
• Retailer A's internal SKU code: A-88213
• Retailer B's internal SKU code: B-5567-CS

If your cross-reference table only maps MFG-00456 to Retailer A, any sales file from Retailer B using B-5567-CS will not join. The product silently drops out of Retailer B's contribution to the brand-level total.

A simple governance check: count of GTINs with zero mapped retailer SKUs, tracked monthly. If that count rises, your consolidated sales report is quietly losing volume. This single metric, sometimes called mapping completeness rate, is a leading indicator of reporting quality that most FMCG data teams should monitor but few do systematically.

If this rate estimate drops from around 98% to 90% month over month (illustrative figures for the exercise, not a benchmark to cite externally), that is roughly 8% of your active product range at risk of being invisible in consolidated reporting, worth investigating before it hits a board deck.

Data qualityData qualityThe degree to which data is fit for purpose: accurate, complete, consistent, timely, valid and unique. Poor quality data undermines analytics, reporting and AI.Voir la définition complète → metrics to track on the product master

A mature FMCG data governancedata governanceData governance is the set of policies, roles, and processes that ensure data is accurate, secure, well-defined, and used responsibly across an organization.Voir la définition complète → function tracks a small set of recurring metrics, usually reviewed monthly by category or master data owners:

• Completeness: percentage of mandatory attributes populated (pack size, net weight, GTIN, launch date). GS1's GDSN validation rules provide a baseline for what counts as mandatory.
• Duplication rate: number of GTINs pointing to what is functionally the same product, often created when regional teams independently register near-identical SKUs.
• Attribute consistency: whether the same GTIN reports identical weight, dimensions, or allergen flags across systems (ERP, retailer portal, GDSN pool). Discrepancies here can trigger real supply chain issues, wrong case counts on pallets, not just reporting noise.
• Hierarchy mapping coverage: the GTIN-to-retailer-SKU completeness metric described above.
• Time to onboard a new SKU: how many days from GTIN creation to full activation across all retailer systems, a governance efficiency measure many manufacturers estimate at two to six weeks depending on retailer complexity.

Practical diagnosis checklist

When two systems disagree on "the same" brand's sales, work through this order:

1. Confirm GTIN-level match first, not description match. Descriptions vary by locale and abbreviation style.

2. Check for pack-size or promotional-pack GTIN variants being excluded or double-counted.

3. Compare category hierarchy depth and node names between the two sources.

4. Check date and calendar alignment (retail weeks versus calendar months, a frequent and separate source of mismatch).

5. Verify currency and unit of measure (cases versus eaches versus units).

For a deeper primer on the identifier standards referenced here, GS1's own explainer is a solid free resource: GS1 GTIN basics.

🎬 [VIDEO: "What is GS1 and How Does it Work?" - youtube.com/@GS1 - a short official explainer on GTIN, barcodes, and the GS1 identifier system underpinning FMCG product master data]

Key Takeaways

• The GTIN, governed by GS1 standards, is the atomic product identifier in FMCG, but a new GTIN is often triggered by pack or promotional changes, which fragments sales history if not tracked carefully.
• No universal category hierarchy exists across retailers. Each grocer builds its own taxonomy, so cross-retailer "category share" numbers require harmonization (which is exactly what providers like Nielsen IQ and Circana sell).
• GDSN and GPC reduce manual data entry and offer shared classification scaffolding, but they do not force retailers into identical hierarchies.
• Track mapping completeness rate, duplication rate, and attribute consistency monthly. These are leading indicators of reporting quality failures, not just IT housekeeping.
• When two systems disagree, diagnose in order: GTIN match, pack/promo variants, hierarchy structure, calendar alignment, unit of measure. Most discrepancies resolve at one of these five checkpoints.